ABSTRACT It is the overall aim of this proposal to dissect the paradoxical mechanism by which the binding of a CAG hairpin converts an otherwise normal MMR complex into a mutational machine. Mammalian cells have evolved sophisticated DNA repair systems to correct mispaired or damaged bases and extrahelical loops. Surprisingly, the eukaryotic mismatch recognition complex, MSH2/MSH3, fails to act as a guardian of the genome and causes CAG expansion, the lethal mutation underlying Huntington's disease (HD) and more than 20 other neurodegenerative diseases. In this proposal, we focus on the two key mutagenic steps that cause the mutation: we will (1) determine why ATP hydrolysis in MSH2-MSH3 fails to signal loop removal, and (2) identify the endonuclease recruited by the MSH2-MSH3-hairpin complex that incorporates the loop into duplex DNA completing expansion. In Aim 1A, we will generate two ?separation-of-function? mutant KI mice for MSH2- MSH3, which bind ATP in each subunit, but lack ATP hydrolytic function in one or the other. If loss of hydrolytic activity in a particular subunit attenuates expansion, then the mutation requires the ATPase activity in that subunit. In Aim 1B, we will solve the crystal structure of MSH2-MSH3 bound to a repair competent (CA)4 loop or to the repair-resistant CAG hairpin. Identified are the structural perturbations in the nucleotide-bound MSH2- MSH3 complex that prevent proper removal of the hairpin loop. In Aim 2, we will identify the canonical and non- canonical endonuclease machinery that facilitates incorporation of the hairpin loop and completes expansion. To identify non-canonical machinery, we will develop technology for site-specific capture of endonucleases ?caught in the act? of incising the loops at the CAG tract during expansion. Inserting a DNA site with CRISPR provides an engineered landing pad for targeting an engineered APEX2 fusion protein. The latter modifies closely located protein partners with biotin, which can be captured on streptavidin plates. We will test how these instructions are misinterpreted for ?in trans? nicking when MSH2-MSH3 is bound to the CAG hairpin. Collectively, the proposed experiments pave the way for small molecule development to restore loop removalby altering the hairpin DNA structure or the protein conformation.